Data Centre Boom Puts Pressure on Power Grids, AI Offers Solutions

A KYRO AI report highlights a paradox: while AI-driven data centres strain power grids, AI technology itself can make grids more reliable and predictive, helping utilities manage rising demand and prevent outages.

The rapid growth of data centres, driven in large part by artificial intelligence, is putting fresh pressure on the power grids, but the same technology could also help utilities make the grid more reliable and better prepared for rising electricity demand, according to a report by KYRO AI.

The report, citing the example of the United States, said data centres used 4 per cent of US electricity two years ago and are on track to use three times that by 2028. The growing demand is creating pressure on an electricity system that was largely designed for a different era, while higher power costs are increasingly affecting households.

AI Offers a Path to a Smarter Grid

The report highlighted that utilities need to move towards smarter and more predictive systems rather than simply continuing to spend on the existing grid. AI is already being used in some parts of the US power sector for this purpose.

Predictive Maintenance in New York

In New York, New York State Electric & Gas is using AI-enabled drones, vehicles and imagery to survey more than 45,000 miles of overhead lines and identify equipment that could fail. The system can identify weak points before equipment breaks down, allowing crews to replace components before they trigger wider outages. The report said work that previously took weeks of manual surveying can now be completed in hours.

Wildfire Mitigation in California

In California, Pacific Gas & Electric is using AI to monitor fire conditions and send automated notifications. According to the report, the utility's combined efforts have brought reportable ignitions down 65 per cent compared with the 2018-2020 average.

California's grid operator, the California Independent System Operator, is also experimenting with a generative AI system called Genie. The system is designed to examine streams of real-time data and identify potential problems before outages spread across the network.

Investment vs. Reliability: A Growing Disconnect

According to Hari Vasudevan, Founder and CEO of KYRO AI, the problem is not simply a lack of investment in the power grid. Utilities have invested heavily in infrastructure, but reliability has not improved at the same pace.

"The problem isn't a lack of investment. Utilities have poured more than a trillion dollars into the grid in the last decade and USD 178 billion in the last year alone. But reliability has barely budged," Vasudevan said. He noted that US electricity customers averaged five and a half hours of power interruption in 2022, compared with about three and a half in 2013. At the same time, electricity rates have increased by more than 30 per cent since 2020, while one in five households now struggles to pay its electricity bill.

Urgent Pressures on the Power Grid

The report said three developments are making the issue more urgent: rising electricity demand from data centres, increasingly severe weather events and the growing use of renewable energy.

The Indian Context: A Parallel Challenge

Recently, the Ministry of Electronics and Information Technology said India's installed data centre capacity has more than quadrupled since 2020 to around 1,575 megawatts (MW), as the country expands its digital infrastructure to support the growing adoption of artificial intelligence (AI).

Installed data centre capacity has increased from 375 MW in 2020 to about 1,575 MW currently, representing a more than fourfold increase over the period, the Ministry said. As per govt data, India's peak electricity demand rose around 12 per cent during April-June as compared to the corresponding period last year and touched an all-time high of 270.8 gigawatts (GW) in May.

So, the report outlined that the growing demand from AI data centres makes upgrading the power grid increasingly important, while AI itself could help utilities make that transition from a reactive system to a more predictive one. (ANI)

(Except for the headline, this story has not been edited by Asianet Newsable English staff and is published from a syndicated feed.)